INDEX

Explanations

descriptions of various locations and events

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iture

-0.81

matter

-0.73

anse

-0.69

ENCE

-0.67

 thereof

-0.65

 Presence

-0.65

aeus

-0.64

fulness

-0.64

brance

-0.64

ence

-0.61

POSITIVE LOGITS

 favorites

0.99

 reasons

0.97

 earliest

0.96

 constants

0.95

 pillars

0.95

 favourites

0.93

 busiest

0.92

 factors

0.90

 quir

0.90

 trademarks

0.89

Activations Density 2.947%

descriptions of various locations and events

No Comments

No Known Activations